Correct folding and assembly of the Rubisco holoenzyme is essential for biological function. For the simpler form II Rubisco this entails the proper folding of the large subunit monomers and assembly of the RbcL dimer, which occurs in the cytosol. Prokaryotic form I Rubisco folding and assembly is also located in the cytosol and rbcL and rbcS are located on the same operon. Red-type form I rbcL and rbcS genes from non-green algae are encoded on the same operon in the chloroplast genome where the proteins are also synthesized (Andrews and Whitney 2003). In contrast, higher plant and green algae green-type form I Rubisco encode the large subunit in the chloroplast genome (encoded as a single copy) with synthesis on the chloroplast ribosome (Blair and Ellis 1973), whereas, the small subunit occurs as a homologous multi-gene family and is encoded in the nuclear genome with synthesis on cytosolic ribosomes (Dobberstein et al. 1977; Hartman and Harpel 1994). The specific role of multiple RbcS proteins in Rubisco is still unknown. RbcS is first synthesized as a precursor molecule in the cytosol with an N-terminal transit peptide for targeting to the chloroplasts
where it is subsequently processed into full-length RbcS. RbcS then assembles on the RbcL8
core, forming the holoenzyme (Bradley et al. 1986; Chua and Schmidt 1978; Goloubinoff et al. 1989b; Hartman and Harpel 1994).
Form II Rubisco and prokaryotic form I Rubisco can be recombinantly expressed in E. coli (Gatenby et al. 1985; Somerville and Somerville 1984). However, no soluble eukaryotic form I Rubisco has been expressed outside of the chloroplast (Gatenby 1984) even if chloroplast chaperonin is also coexpressed in E. coli (Cloney et al. 1993). Early experiments of Rubisco revealed that a binding protein (Cpn60) was associated with newly synthesized Rubisco in chloroplasts (Barraclough and Ellis 1980; Bloom et al. 1983; Roy et al. 1988), yet at that time the exact role of this protein was not known. An important breakthrough in understanding a critical step in Rubisco biogenesis arose when it was demonstrated that a certain subset of chaperones, the chaperonins, are required to obtain assembled Rubisco folding in a bacterial host system (Goloubinoff et al. 1989b) and for in vitro refolding of form II Rubisco (Goloubinoff et al. 1989a). The bacterial DnaK/DnaJ/GrpE chaperone system also aids in folding of recombinantly expressed Rubisco (Checa and Viale 1997), but the chaperonin system was found to be essential for the folding and possibly the assembly of Rubisco large subunits. This advance was due to the realization that the bacterial chaperonin system, GroEL and GroES, are evolutionary homologues of the chloroplast chaperonin system, Cpn60 and Cpn21 (Hemmingsen et al. 1988). The action of chaperonin in assisting RbcL folding is essential for Rubisco assembly. Once the RbcL8 core of form I Rubisco is
formed, RbcS binds and assembles to give the active holoenzyme in an ATP-independent manner (Gatenby and Ellis 1990; Hartman and Harpel 1994).
To date higher plant Rubisco has failed to reconstitute upon expression in E. coli. However, the reason for this remains ambiguous especially since prokaryotic and eukaryotic form I Rubisco from different species exhibit high sequence identity and structural similarity. One possible explanation is that eukaryotic Rubisco undergoes co- and post-translational modifications which are not fulfilled upon expression in E. coli e.g., deformylation, acetylation, methylation, and N-terminal proteolytic processing of the holoenzyme (Houtz et al. 2008). This possibility seems less likely because many of these modifications occur after assembly and influence overall stability or activity and not necessarily assembly. It is more likely that specific folding or assembly factors are missing in E. coli that occur in the chloroplast which are necessary for holoenzyme formation. Large protein complexes consisting of multiple subunits have been shown to require assistance from chaperones not only for folding but for correct assembly. For example accessory assembly factors are utilized
for the assembly of the nucleosome (Laskey et al. 1978) and the proteasome (Witt et al. 2000). Indeed, for Rubisco, one such assembly factor, RbcX, has been characterized and shown to be involved in cyanobacterial form I Rubisco assembly (Li and Tabita 1997; Liu et al. 2010; Saschenbrecker et al. 2007; Tarnawski et al. 2008).
2.3.4.4
RbcX
Figure 2.22 Amino acid sequence alignment of RbcX2 from prokaryotes and eukaryotes.
(A) Amino acid sequence alignment of RbcX2 from Syn7002 compared to eukaryotic species without putative-transit
peptide: C. reinhardtii (Cr), A. thaliana (At), Zea mays (Zm) without putative transit peptides. Isoforms of eukaryotic RbcX2
are indicated as RbcXI and RbcXII. Conserved and important regions are indicated with a star and the numbering above the
sequences is according to Syn7002. (B) Amino acid alignment of RbcX2 from cyanobacteria, labeling as in A. Conserved
residues are shown in red and residues found in two of the sequences are shown in blue. Black indicates no homology. Alignment created using MultAlin.
The discovery of the small assembly chaperone, RbcX, revealed that besides the chaperonin system there is another protein involved in Rubisco biogenesis (Li and Tabita 1997). Homologues of rbcX have been found in genomes of species encoding form IB Rubisco including cyanobacteria, green algae, and higher plants. In some cyanobacteria species (Anabaena sp. PCC7120, Ana7120; Anabaena sp. CA, AnaCA; Synechococcus sp. PCC7002, Syn7002) the rbcX gene is encoded on the rbc operon between the rbcL and rbcS
genes, and deletion of the rbcX gene region hinders the production of soluble Rubisco (Larimer and Soper 1993; Onizuka et al. 2004). In Synechococcus sp. PCC6301 (Syn6301) the rbcX gene is located downstream of the rbcL and rbcS operon. Syn6301 Rubisco can assemble in E. coli without expression of the rbcX gene; however coexpression of rbcX
enhances the production of assembled Rubisco (Saschenbrecker et al. 2007). The rbcX gene is *Y17 *Y20 *Q29*E32*N34 *R70
*Y17 *Y20 *Q29 *E32*N34 *R70
A
encoded in the nuclear genome of green algae and higher plants with at least two isoforms of the rbcX gene present in the genome of higher plants (Figure 2.22 A).
RbcX is a homodimer of ~15 kDa subunits consisting of four α-helices aligned in an anti-parallel fashion along the α4 helix held together by uncharged/hydrophobic interactions (Figure 2.23). RbcX2 functions as a stabilizer of folded RbcL. The RbcX2 protein recognizes a
highly conserved C-terminal sequence of the large subunit of form IA and form IB Rubisco, termed the C-terminal recognition motif: EIKFEFD (Figure 2.24). After RbcL interacts with and is folded by the chaperonin system, RbcX2 stabilizes the folded RbcL and then assists in
the formation of the RbcL8 core. RbcX2 does not remain a part of the final holoenzyme
complex, instead RbcX2 has a dynamic relationship with RbcL. RbcX2 is displaced by RbcS
binding, probably due to a conformational change in the large subunits (Liu et al. 2010; Saschenbrecker et al. 2007). However, the exact mechanism by which RbcS binding displaces RbcX2 from the RbcL8 core is still unknown.
Figure 2.23 Crystal structure of Syn7002-RbcX2 and RbcX2 conserved surface regions.
(A) Structure of the RbcX monomer. N- to C-terminal is indicated by a gradient of cold to warm colors, α-helices 1-4 are labeled along with numbering of specific amino acids along the chain. (B) RbcX dimer structure. Subunits are depicted in yellow and light blue; N- and C-termini are labeled. (C) Amino
acid surface conservation in RbcX2. 151
cyanobacterial RbcX2 sequences were
aligned and the similarity score plotted
onto the exposed surface on RbcX2
using the PFAM database. Highly conserved residues are shown in magenta and variable residues are shown in cyan. Conserved surface residues are labeled. Adapted from: (Saschenbrecker et al. 2007)
Similarity scores from an alignment of 151 RbcX2 sequences from cyanobacteria
revealed that there are two highly conserved regions on RbcX2: the 5.4 Å wide central
hydrophobic crevice and the peripheral polar surfaces at the corners of the dimer (Figure 2.23). From mutational studies, it was found that both of these regions are critical for RbcX2
function. The central crevice is important for the production of soluble RbcL by binding to the extended C-terminal recognition motif. Two phenylalanines (462 and 464, Figure 2.24 C: numbering according to Syn7002 Rubisco) extend into the hydrophobic groove of RbcX2.
The peripheral surface is important for efficient RbcL8 core complex formation. Importantly,
A
B
residues that are necessary for cyanobacterial RbcX2 function are also highly conserved in the
sequences of higher plant RbcX2 i.e., Y17, Y20, Q29, E32, N34, R70 (Figure 2.22). The
dynamic relationship of the RbcL/RbcX complex is critical for the formation of the holoenzyme. Comparison of RbcX2 from cyanobacteria has shown that they share only about
50% sequence identity (Figure 2.22 B), however crystal structures of various prokaryotic RbcX2 proteins have indicated high similarity of tertiary and quaternary structures
(Saschenbrecker et al. 2007; Tanaka et al. 2007).
Figure 2.24 Binding of RbcX2 to RbcL C-terminal recognition motif.
(A) Alignment of the C-terminal residues of RbcL from cyanobacteria and higher plant species as indicated using MultAlin. Swiss-Prot accession numbers are shown in brackets. Red coloring indicates high (greater than 90%) consensus level and blue coloring indicates low (less than 50% consensus level). Conserved RbcL C-terminal recognition motif is boxed (B)
Surface representation of RbcX2 crystal structure (individual subunits shown in white and blue) in complex with the peptide
EIKFEFD in stick representation. N- and C- termini of the peptide are shown. (C) Close-up of the boxed-in area in B, the
interactions of the EIKFEFD peptide with the groove region of RbcX2 are highlighted. Hydrogen bonds are indicated with
dashed lines. Residues of RbcX2 that participate in peptide binding are shown in stick representation below the transparent
surface and colored white or yellow for each subunit. The phenylalanines of the C-terminal peptide that bind the hydrophobic
pockets in the RbcX2 groove are also labeled. Adapted from: (Saschenbrecker et al. 2007)
It was observed that when RbcL and RbcX2 from different species are used together, a
stable, but nonetheless dead-end complex was formed (e.g., Syn7002-RbcL with AnaCA- RbcX2). The affinity of AnaCA-RbcX2 to the C-terminal recognition motif of Syn7002-RbcL
is extremely high (KD: 5 µM) and Syn7002-RbcS is not able to replace AnaCA-RbcX2,
hindering holoenzyme formation (Saschenbrecker et al. 2007). RbcX2 is a substrate specific
assembly chaperone; thus, these previous results indicate that the RbcL and RbcX2 dynamic
interaction evolved in a way where displacement of RbcX2 by RbcS is optimal for proteins
from the same species. The dynamic nature of the RbcX/RbcL complex has hindered attempts to characterize the regions of RbcL that interact with the peripheral regions of RbcX2. In the
current model of cyanobacterial form I Rubisco folding and assembly, the large subunit
A
undergoes folding in the GroEL/ES cage. Upon release of the large subunit from the chaperonin, RbcX2 binds the C-terminal recognition motif and promotes the formation of
RbcL8 complexes. Binding of RbcS displaces RbcX2 thereby creating the functional
holoenzyme (Figure 2.25) (Liu et al. 2010; Saschenbrecker et al. 2007).
Figure 2.25 Model of form I Rubisco folding and assembly.
1. RbcX2 interaction occurs
downstream of GroEL/ES action. 2.
Binding of RbcX2 stabilizes folded
RbcL at dimer or monomer level. 3. Subsequently followed by formation
of the RbcL8 core. 4. The
RbcL/RbcX interaction is dynamic,
and RbcS binding to the RbcL8 core
results in the displacement of bound
RbcX2 resulting in holoenzyme
formation.
Adapted from: (Saschenbrecker et al. 2007)
Form II Rubisco was the only Rubisco to be successfully reconstituted in vitro from the denatured state using the GroEL/ES system (Goloubinoff et al. 1989a). However, more recently, form I Rubisco from Syn6301 cyanobacteria was successfully reconstituted in vitro
(Liu et al. 2010). The GroEL/ES system and the assembly chaperone RbcX2 are the two key
chaperones used for the in vitro reconstitution. It was demonstrated that when RbcX2 was not
present in the assay, Syn6301-RbcL was not released in an assembly competent manner, and hence rebound to GroEL. Furthermore, addition of cognate RbcX2 to the assay also failed to
produce assembled RbcL8 core complexes. Folded RbcL could only be removed from GroEL
if the high affinity heterologous RbcX2 was added to the system. The high affinity RbcX2
could compete with GroEL for binding of RbcL and push the equilibrium to RbcL/RbcX assembly competent intermediates. As noted above, high affinity heterologous RbcX2 inhibits
proper formation of holoenzyme in that it is not displaced by RbcS creating a dead-end complex; in the in vitro assay it was observed that increasing concentrations of high affinity RbcX2 was detrimental to catalytic activity. However, addition of the RbcL C-terminal
recognition motif peptide, after assembly of the RbcL/RbcX complex aided in removal of RbcX2 from the RbcL8 core, and as observed on Native-PAGE the holoenzyme, RbcL8S8,
could form (Liu et al. 2010). The development of an in vitro reconstitution assay for form I Rubisco is a vital advancement for future efforts to engineer a ‘better’ Rubisco enzyme by creating a fast and efficient method for screening many Rubisco mutants.